UMBC High Performance Computing Facility
Please note that this page is under construction. We are documenting the
240-node cluster maya that will be available after Summer 2014.
Currently, the 84-node cluster tara still operates independently,
until it becomes part of maya at the end of Summer 2014.
Please see the 2013 Resources Pages under the Resources tab for tara information.
Toward a better understanding of the vertical structure of
marine boundary layer clouds using MODIS observations and large eddy
Zhibo Zhang, Department of Physics
In situ observations indicate the existence of significant and
highly variable microphysical vertical structures in marine boundary
layer (MBL) clouds due to the combined effects of condensational growth,
entrainment mixing, collision-coalescence, and sedimentation.
An understanding of this vertical structure (e.g., liquid water
content and cloud effective particle radius) is important for a variety
of reasons. First, cloud vertical structure reflects the interaction and
competition of cloud processes in shaping the macro- and microphysical
properties of MBL clouds. Second, an understanding of cloud vertical
structure is crucial for assessing the uncertainties caused by the
homogeneous cloud assumption in cloud remote sensing. In addition, it is
also an essential step toward possible MODIS-based retrieval cloud
droplet number concentration (CDNC) and drizzle, two key parameters for
studying aerosol indirect effects.
Despite its importance, the vertical structures of MBL clouds on a
global scale remains largely unexplored. Although space-borne active
sensors, such as CALIOP and CloudSat, are designed to resolve cloud
vertical structure, some challenging issues and inherent limitations
hinder their application to MBL clouds. The MODIS instruments on both
Terra and Aqua have three shortwave infrared (SWIR) bands centered at
about 1.6, 2.1 and 3.7 μm. These SWIR bands have different cloud
penetration depths, and therefore carry information about different
layers in a cloud. Previous studies have shown that the combination of
these MODIS bands can be used for studying cloud vertical structure.
However, several factors, such as 3-D radiative effects, algorithm
issues and instrument radiometric accuracy complicate the information
contained in these bands. Motivated by the need for improved knowledge
of MBL cloud vertical structure and the recognized need to understand
potential biases in MODIS retrievals, we propose a comprehensive
modeling investigation. We intend to first evaluate the cloud vertical
structure information content in MODIS-like observations from simulated
MBL large-eddy simulation (LES) cloud fields and 3-D radiative transfer
models. Based on results from these simulations, we will explore the
capabilities, as well as limitations, of MODIS observations for studying
temporal and spatial variations in the vertical structure of MBL clouds.
Our specific goals are as follows:
Integrate LES model, 3-D radiative transfer model and a MODIS
retrieval simulator into an analysis tool for analyzing MODIS-like MBL
cloud property retrievals;
Achieve a better understanding of the information content in MODIS
observations on the vertical structure of MBL clouds by applying
MODIS-like retrievals to LES cloud simulations;
Explore the theoretical basis for potential future MODIS cloud
algorithms, including cloud adiabatic index, cloud droplet number
concentration (CDNC) and drizzle flags for MBL clouds;
Improve our understanding of the uncertainties in current MODIS
operational cloud products;
Study vertical cloud structure as revealed by MODIS and other
A-train sensors over key MBL cloud regimes.